查看更多>>摘要：At about this time last year, one of us (MRC) edited a manuscript submitted by systematic ichthyologists Randall D. Mooi and Anthony C. Gill to Zootaxa entitled "Phylogenies without synapomorphies—a crisis in fish systematics: time to show some character" (Mooi & Gill, 2010; hereafter M&G). A preview of this manuscript had been presented at an American Society of Ichthyologists and Herpetologists (ASIH) meeting in 2008, and caused quite a stir among many in attendance. The strong reaction was a response to what was perceived as unfair criticism of the molecular paradigm in the phylogeny of fishes, particularly of percomorphs, the core of their presentation. Before publishing M&G in Zootaxa, the authors were advised that a similarly strong reactionwould probably occur. They did not flinch, and hence M&G was published (after peer review). Needless to say, M&G got strong feedback, either in a negative or positive vein, from molecular and morphological workers, respectively. MRC even received a critical e-mail from one molecular worker accusing M&G of superficiality and Zootaxa of employing "low" standards of scholarship.
EDWARD O. WILEYPROSANTA CHAKRABARTYMATTHEW T. CRAIGMATTHEW P. DAVIS...
10页
查看更多>>摘要：In a recently published commentary, Mooi & Gill asserted that there is a crisis brewing in systematic ichthyology caused by a failure of investigators to apply the basic tenets of outgroup comparison to recover clades based solely on shared apomorphiccharacters. The result, they claim, is that many recent analyses disregard real synapomorphies and discover clades by phenetic rather than phylogenetic principles. We take the opportunity to refute this claim and assert that matrix-based analyses, whether parametric or nonparametric, satisfy the basic tenets of Hennig's methods, resulting in monophyletic groups confirmed by synapomorphies.
EDWARD O. WILEYPROSANTA CHAKRABARTYMATTHEW T. CRAIGMATTHEW P. DAVIS...
10页
查看更多>>摘要：In a recently published commentary, Mooi & Gill asserted that there is a crisis brewing in systematic ichthyology caused by a failure of investigators to apply the basic tenets of outgroup comparison to recover clades based solely on shared apomorphiccharacters. The result, they claim, is that many recent analyses disregard real synapomorphies and discover clades by phenetic rather than phylogenetic principles. We take the opportunity to refute this claim and assert that matrix-based analyses, whether parametric or nonparametric, satisfy the basic tenets of Hennig's methods, resulting in monophyletic groups confirmed by synapomorphies.
查看更多>>摘要：The assertion that phylogenetic inference algorithms are not authoritarian because results are repeatable, predictable and freely available misses the point that the authority resides in underlying algorithm models that are not cladistic. We show thatoptimization procedures can group using symplesiomorphy and that optimization is not always equivalent to cladistic argumentation. Because parsimony and Bayesian algorithms can obtain the same answer from the same data set is not evidence that they areHennigian; examples exist where these methods do not provide the same result from the same data. Using 'reversals' as evidence in systematics is problematic because the question, "Reversal to what?" has no straightforward answer. This confusion can be eliminated by recognizing that homologues are the parts of organisms and homologies are the relationships between the parts, and that the latter is a hierarchical concept rather than transformational. We clarify that Hennig's auxiliary principle pertains to potential synapomorphy, meaning for molecular work that it is the presence of a particular derived nucleotide that is shared in a given position of aligned sequences of two or more taxa that should be considered homologous until proven otherwise, not simply the alignments themselves. We reiterate that not all data are evidence and we specifically reject homoplasy as a source of 'evidence' for systematics. We further reject the view that conflict among data should be resolved through methodology. It is the data that should be our primary focus, as it is our attempts to identify and clarify homologues worthy of suggesting relationships (homology) that are primary in systematics.
查看更多>>摘要：In short, the charges leveled against us by Wiley et al. (2011) do not stand up to scrutiny. We had no ulterior motive in bringing what we consider to be legitimate concerns to the attention of our research community. We hope that the real issues—asdetailed in our paper (Mooi & Gill 2010)—can now be discussed without further distraction. Wiley et al. (2011) begin their critique of our paper (Mooi & Gill, 2010) with an assertion: "we need to make it clear that the foundation of their arguments rests not on scientific rigor, but rather on opinions about the re-classification of fishes using molecular data.
MATTHEW P. DAVISEDWARD O. WILEYPROSANTA CHAKRABARTYMATTHEW T. CRAIG...
5页
查看更多>>摘要：We thank the editor of this special series in allowing us a short response to Mooi, Williams and Gill (2011). The space allocated for this rebuttal is small, so we address only the basic points, organized by their major headings.Mooi et al. again suggest that finding trees under various universally-used optimality criteria is "authority." We reject this argument. Their alternative of picking "good" characters that happen to evolve at a certain rate is equally authoritarian (indeed, it is easy to characterize just about any method as "authoritarian" because, well, it's a method; the only thing fully immune from the charge of "authoritarianism" is doing nothing). To explain: the usual manner in which characters are selected for morphological analysis follows three basic steps. (1) Characters that display random variation or as much variation within taxa as between taxa in an analysis are typically discarded. (2) Characters that show no variation between taxa in the study are discarded. (3) Characters that do not meet the above criteria but which are shared between two to N-l taxa in the study are analyzed. This amounts to a rate model of evolution. We don't see any particular problem with adopting such models, but investigatorsshould recognize them as models with their own optimality criteria. We might term it the "screen data for usefulness model." Mooi et al. argue in favor of a particular brand of this model, Three-Taxon Analysis (3ta), in which characters that change only once (i.e., without reversals) are chosen. We explain below why taking this course would be restrictive and misguided.
查看更多>>摘要：Mooi & Gill (2010; hereafter M&G) presented a discussion of current methods in molecular systematics and contended that methodologies employed in the analysis of molecular data do not identify homology or synapomorphy, are phenetic in nature, and haveresulted in a categorical movement "away from an intimate understanding of character distribution, homology, and the meaning of evidence" (M&G, p. 26). If they are correct in their assessment, systematic ichthyology (and systematic research in general)truly is in a state of crisis, and many of the advancements that have been made in the last two to three decades have, in reality, been steps backward. As with any discipline, papers pointing out the dangers of dogmatic thinking and the perils of uncritical analysis are often welcome "wake up calls" and serve a genuine and necessary purpose in scientific literature. Unfortunately, the paper by M&G appears nothing more than a rehashing of long-settled debates with not-so-subtle hints of deeply rooted animosity towards a few specific researchers. Their paper not only misses their intended mark (presumably to ensure the scientific integrity of systematic ichthyology), but also demonstrates a clear misunderstanding of how molecular data are accumulated, analyzed and interpreted.
查看更多>>摘要：Mooi & Gill (2010; hereafter M&G) presented a discussion of current methods in molecular systematics and contended that methodologies employed in the analysis of molecular data do not identify homology or synapomorphy, are phenetic in nature, and haveresulted in a categorical movement "away from an intimate understanding of character distribution, homology, and the meaning of evidence" (M&G, p. 26). If they are correct in their assessment, systematic ichthyology (and systematic research in general)truly is in a state of crisis, and many of the advancements that have been made in the last two to three decades have, in reality, been steps backward. As with any discipline, papers pointing out the dangers of dogmatic thinking and the perils of uncritical analysis are often welcome "wake up calls" and serve a genuine and necessary purpose in scientific literature. Unfortunately, the paper by M&G appears nothing more than a rehashing of long-settled debates with not-so-subtle hints of deeply rooted animosity towards a few specific researchers. Their paper not only misses their intended mark (presumably to ensure the scientific integrity of systematic ichthyology), but also demonstrates a clear misunderstanding of how molecular data are accumulated, analyzed and interpreted.
查看更多>>摘要：We were reticent to respond to Craig (2011), for worry that this might lead to an ongoing, futile exchange on who does and who does not understand homology and the workings of systematics. No reply, too, would be a reply of a sort. However, Craig's missive provides an opportunity to explore additional aspects of the issues/crises at hand. Before we do so, we would point out that in Mooi & Gill (2010a) we did not say that molecules cannot provide homology (as Craig implied), but instead argued that current popular methods are phenetic and do not discover this homology. Numerical methods do not differentiate homology from homoplasy in morphology consistently, either, and thus are a disservice to the molecular and morphological evidence; the phenograms produced do not necessarily appropriately or accurately reflect the relationships the data can provide.
查看更多>>摘要：Mooi & Gill (2010) have made a number of criticisms of statistical approaches to the phylogenetic analysis of molecular data as it is currently practiced. There are many different uses for molecular phylogenies, and for most of them statistical methods are entirely appropriate, but for taxonomic purposes the way that these methods have been used is questionable. In these cases it is necessary to introduce an extra step into the analysis - exploration of character conflict. Existing methods for exploring character conflict in molecular data such as spectral analysis, phylogenetic networks, likelihood mapping and sliding window analyses are briefly reviewed, but there is also a need for development of new tools to facilitate the analysis of large datasets. Incorporation of previous phylogenies as priors in Bayesian analyses could help to provide taxonomic stability, while still leaving room for new data to alter these conclusions if they contain sufficiently strong phylogenetic signal. Molecular phylogeneticists should make a clearer distinction between the different uses to which their phylogenies are put; methods suitable in one context may not be appropriate in others.
NSTL